## ABSTRACT

Fresh water is considered to be one of the most critical strategic and economic resources in

coming decades. However, leakage often makes up a substantial fraction, frequently more than 30%, of the water supplied. The objective of this paper is to present an improved formulation of the hydraulic network equations that incorporates both fixed and pressure dependent leakage, Particular effort has been made to assess the magnitude of the different terms in the equations, including leakage inertia and head loss. A sensitivity analysis was also conducted as part of this study. Two formulations were derived from the Navier-Stokes equations and subsequently solved: the first considers constant leakage in the water distribution network and leads to an ordinary differential equation. This equation is Solved Using a generalised Runge-Kutta method. In the second formulation, pressure dependant leakage is taken into account. A partial differential equation is then solved using the finite difference method.

A data-driven strategy for conducting water distribution network pipe roughness calibration is introduced and explained. The approach is based on solving the inverse calibration problem by using measured head data from only selected locations in a network and then applying an evolutionary search strategy known.as Evolutionary Polynomial Regression (EPR) in order to encapsulate network behaviour. EPR combines genetic programming, for symbolic regression, with genetic algorithms, for parameter estimation, in order to return straightforward polynomial formulas, in this case relating hydraulic conductivity with nodal head. The intention is that these equations could serve as either roughness approximations for some network pipes or help to identify which pipes and nodes may be of interest for closer examination in conventional calibration field tests, while also remaining easily comprehensible for system operators. The concept is illustrated with a case study based on a proposed design for a real system in Puglia, Italy.

This article describes an algorithm for the pressure driven analysis of water distribution networks. Such an approach involves two main issues: the determination of a sound discharge-pressure relationship and robustness of the numerical algorithm used to solve the system of equations goveming network hydraulics, To this end, the well-known Wagner relationship between nodal flow and pressure is employed and an initial solution of the system equations is assumed. These assumptions eschew complex discharge-pressure relationships, avoiding numerical problems in regions where the Wagner function presents a discontinuity and is not differentiable, while the adoption of a proper initial search point allows for good convergence towards the solution.

The current climate of interest in water-environmental issues has lead to increased use of advanced computer-based simulation techniques in the evaluation and improvement of water management Systems. This paper reviews the usage history of Computational Fluid Dynamics (CFD) techniques by a supplier of technologies for urban water management, leading to the development of insights and guidance on the adoption of such tools, presented in the context of a number of practical case studies. The paper finds that CFD methods offer genuine opportunities and can yield direct tangible benefits. However, this is reliant on organisations having a realistic understanding of what such approaches can offer, their limitations, and also having an appreciation of the long-term commitment that is required to achieve maximum value.

System Dynamics Modelling is a methodology for studying and managing complex feedback systems, typically used when formal analytical models do not exist, but system simulation can be developed by linking a number of feedback mechanisms. They are particularly useful for building, understanding and presenting models to non-engineers. In this paper the procedure for developing conceptual and System Dynamic Modelling in participatory interdisciplinary context is presented, followed by a specific application case study, i.e. the conceptual model and SDM for the water system of Kremikovtzi in Sofia, Bulgaria.

A water distribution system which consists of a network and a pumping station is considered. The energy characteristics of network and pumps are analysed in interaction. The global energy performance and hydraulic power capacity of the distribution system as a whole are considered. The concepts of the hydraulic efficiency of the distribution system are defined. On the basis of the above the maximum energy efficiency of the distribution system is determined.

The. original algorithm development of direct calculation pump speed by Todini, Tryby, Wu and Walski (2010) was to maintain a target hydraulic head while this paper extends the algorithm and its implementation for direct calculation of pump speed to deliver the fixed flow via a variable speed pump. The fixed flow pumping is desirable in practical conditions where a fixed amount of flow needs to be supplied to the system to meet demand and/or fill a storage tank. The implementation shows that the extended algorithm is very straightforward based upon the original development. An example system is tested to verify the enhanced development and also to demonstrate the application of modelling the fixed flow variable speed pumps.

The traditional optimization paradigm used in system design focuses on network cost minimization and the maintenance of nodal pressures by relying on a demand-driven simulation of the network. This article describes a more realistic optimal design approach in which simulations are conducted with a pressure-driven model in order to assess unsupplied network demand and actual leakage flow rates. Reliability, within the context of unsatisfied demand and leakage, serves as an additional objective criterion in the optimization formulation. The assumption of a maximum network deterioration level is revisited and shown to be useful for improving reliability in the face of escalating leakage.

A novel approach to determine optimal sampling locations under the parameter uncertainty in a Water Distribution System (WDS) for the purpose of its hydraulic model calibration is presented. The problem is formulated as a multi-objective optimisation problem under calibration parameter uncertainty. The objectives are to maximise the calibrated model accuracy and to minimise the number of sampling devices as a surrogate of sampling design cost. Model accuracy is defined as the average of normalised traces of model prediction covariance matrices, each of which is constructed from a randomly generated sample of calibration parameter values. To resolve the computational time issue, the optimisation problem is solved using a Multi-Objective Genetic Algorithm and Adaptive Neural Networks (MOGA-ANN). The results show that significant computational savings can be achieved by using MOGA-ANN compared to the Monte Carlo Simulation (MCS) model or the GA model based on all full fitness evaluations without significant decrease in the final solution accuracy.

In the past, the pimp regime, of sewer pressure mains was based on the cycle frequency of the pump in combination with the storage capacity of the wet well of the pumping station. Modem pump units have higher pump cycle frequencies, enabling a reduction in size of the wet well of the pumping station. In view of construction costs of the pumping station, this trend of down-sizing is desired. A second development that took place in the Netherlands at the same time was the Construction of mildly sloped, rather than steeply sloped pipe sections underneath obstacles such as railways and channels, due to anew drilling technique. A problem arises for these pressure mains when gas is present in the system. The long travel time through a mildly Sloped pipe section, in combination with Short pump operation, hinders the gas bubbles to travel through the sloped part. After pump stop, the gas bubbles return to the high point in the system, eventually a causing capacity reduction, The pumping station and its pump operating condition must be designed in close relation with the pipeline system. This paper describes and discusses new design aspects and proposes guidelines that have to be taken into account to avoid gas problems when designing a wastewater pressure main.

The macroscopic model of water pressure at measurement nodes can estimate the value of measurement nodes according to the change of pressure and flow at each pump station by the method of least squares regression. However, there is strong multi-correlation between multi-dependent variables (the pressure in each measurement node) and multi-independent variables (the pressure and flow of each pump station), which results in harmful effect on variables estimation and robust of the model with the traditional method of least squares regression. However, Partial Least Squares (PLS) regression is an important statistical modeling method, which synthesizes principal component analysis, representative correlation analysis and multiple linear regression. Furthermore, the method can deal with the problem of the bad effect of multi-correlation among variables in such model by analyzing and screening the data, The method has been tested on estimating the value of pressure measurement nodes in a city of China, and the result has an agreement with the measurement data.

The accuracy of a Water Distribution System (WDS) model depends on how well it has been calibrated. Analysis of measurements showed that measured pressures often contain systematic errors. Another type of errors is inaccurate estimation of the elevations of junctions* which will lead to differences between measured and modeled pressures. In this paper we propose to use pressure differentials for the calibration of operational water distribution system containing thousands of pipes with different age. The pump speed is regulated in a way that pressure in the junction next to pump, is constant. Pressures in other junctions depend on junction elevation and head loss in pipes. Head loss, in turn, depends on water consumption and roughness of pipe. Pressure differentials in a junction at times with high and low consumption do not contain systematic errors of loggers and errors due to inaccurate elevation of junction. It is shown also that this approach eliminates significantly the influence of leakages as well.

The current paper presents the results of the hydraulic transient analysis of a hydropower system with multiple uses in Madeira Island (Portugal) obtained both during the design stage and during a field data collection programme, The case study consists of a conventional hydropower system with two hydropower plants (Socorridos and St. Quitéria), that has been converted into a pump storage hydropower system by means of the construction of two storage tanks and a pumping station. Given the new operating conditions at the pumping station, a hydraulic transient study has been carried out and a surge protection device was specified. Transient tests Were carried out in the new system for different operating conditions. Neither negative nor over pressures occurred suggesting that the system was adequately protected against water hammer. However, calculated transient pressures did not fit with collected physical data. Conclusions are drawn about the main sources of uncertainty in the numerical simulations carried out.